Longitudinal section of the whole simulated track, the full 7.6 mm domain. Dark is the fusion boundary that has already set (this is what a metallographic cut shows); orange is the metal that is fully liquid at this moment (liquid fraction 1). Behind the beam the track falls into three stretches, and their boundaries are marked. The penetration depth fluctuates irregularly along the track and still rises with x: within the 250 ms simulated, the pool is still deepening and lengthening, which is not long enough to observe a steady pool behaviour.
Vertical exaggeration —×. Drag the slider to watch the pool advance along the track; the dark fusion record freezes behind it. The blue band is the extent covered by the beam, D4σ = 2.00 mm (r₀ = 707 µm).
The orange region encloses fully liquid metal on the symmetry plane y = 0: T ≥ 1729.1 K, the liquidus of the solver's solidification model. The solver sets the liquid fraction linearly in temperature between the solidus 1661.3 K and the liquidus 1729.1 K (the phase lines of the nominal bead composition f = 0.3232, used for every composition), so the liquid fraction is 1 inside the orange region; the 1661.3–1729.1 K layer just outside it is the mushy zone, with a liquid fraction between 0 and 1. The dark fusion boundary is read from the peak temperature 1771 K (the liquidus of 4140, decision D5), which is a different quantity.
The simulated outline is the mean of the 40 sections that are fully solid at 250 ms (x = 3.01–3.95 mm): the fusion boundary is the iso-surface peak temperature = 1771 K, and the bead surface is the metal height column by column (at 250 ms the once-molten metal in these sections is still at 949–1613 K, median 1147 K, and the surface still includes the thermal lift of the substrate). The measurement is the etched cross-section of the same condition (0822 curve, SHA256 3e4a2b03fc1d…). Both are aligned with the original substrate surface at z = 0 and the pool centre at y = 0, at the same scale horizontally and vertically.
| Quantity | Simulation | Measured | Difference | Simulation basis |
|---|
Follow one fixed cross-section of the track and start the clock when the beam centre is over it; distances convert to time with the scan speed 18.333 mm/s. The beam is 2.00 mm wide by D4σ, and its trailing edge (1.00 mm behind the centre) leaves the section only at +54.5 ms.
The beam centre has passed, but the section is still under the rear half of the beam (it leaves the beam only at +54.5 ms) and keeps being heated; the melting front advances downwards and sideways. The fusion boundary read from the peak temperature 1771 K reaches its final position about 42 ms after the beam centre has passed (decision D27).
The fusion boundary no longer moves outwards; this is the outline a metallographic cut shows. The section still holds liquid metal, but the liquid only shrinks; until it solidifies, the section still receives CCN carried back by the flow from the pool ahead.
All liquid has gone. In the 250 ms snapshot the tail of the liquid trails the beam centre by 2.11 mm; this interval grows as the pool lengthens: 104 ms at 200 ms, 111 ms at 225 ms, 115.2 ms at 250 ms.
The 96 set sections at x = 3.01–5.29 mm at 250 ms (the beam centre passed them more than 42 ms earlier, so their fusion geometry no longer changes). Their penetration depth fluctuates from section to section and still rises along x, so the table gives the mean and the range, not a steady value.
| Quantity | Simulation mean | Simulation range | Measured | Difference of means |
|---|---|---|---|---|
| Depth D / µm | 144.1 | 121.3 – 165.4 | 181.5 | -20.6% |
| Width W / µm | 1112 | 1074 – 1135 | 974.2 | +14.2% |
| Aspect ratio D/W | 0.1295 | 0.1863 | -30.5% | |
| Fusion area / µm² | 108 997 | 117 175 | -7.0% |
The simulated bead is compared with the measured bead area 55 958 µm² through the CoCrNi mass per unit track length ÷ 7900 kg/m³.
The 40 fully solid sections at 250 ms (x = 3.01–3.95 mm): the mean bead area converted from mass is 67 771 µm² (65 090 – 70 107), 116%–125% of the measurement (mean +21.1%). The excess comes from the pool still lengthening: between 225 and 250 ms the tail of the liquid advances at only 15.4 mm/s, slower than the beam's 18.333 mm/s, so each solidified stretch of track receives more CoCrNi than feed rate ÷ scan speed. Once the pool stops lengthening, the two should be equal.
The metal area at z > 0, Σα, cannot be used for the comparison. Over the same sections its mean is 99 811 µm², and it contains two terms unrelated to deposition: first, the substrate expands when heated and its surface lifts — sections at x < 0.6 mm (never molten, never reached by powder) also show 11 814 and 12 237 µm² at 225 ms and 250 ms; second, the section keeps cooling and contracting after solidification, and this area shrinks with it (see the "Melt Pool Mixing Timeline" page).
Over the set stretch x = 3.01–5.29 mm the depth rises at +7.1 µm/mm (dashed line: linear fit); the standard deviation of the sections about the fit is 9.1 µm. The grey band is the stretch not yet set. The red dashed line is the measured 181.5 µm.
The deepest point of the fully liquid region (metal at T ≥ 1729.1 K on the symmetry plane) is always behind the beam centre: 0.22 mm at 25 ms, 0.72 mm at 250 ms (0.22–0.82 mm over 25–250 ms).